Re: ichwd0: unable to reserve GCS registers

On Wednesday, August 03, 2011 2:19:24 pm Test Rat wrote:
> John Baldwin <jhb_at_freebsd.org> writes:
> 
> > On Wednesday, August 03, 2011 1:23:29 pm Test Rat wrote:
> >> John Baldwin <jhb_at_freebsd.org> writes:
> >> 
> >> > On Wednesday, August 03, 2011 4:49:24 am Test Rat wrote:
> >> >> Doug Barton <dougb_at_FreeBSD.org> writes:
> >> >> 
> >> >> > On 08/02/2011 15:06, John Baldwin wrote:
> >> >> >> On Saturday, July 30, 2011 2:49:52 am Andriy Gapon wrote:
> >> >> >>> on 19/07/2011 18:16 John Baldwin said the following:
> >> >> >>>> Hmm, can you get devinfo -r output from a working kernel with 
ichwd 
> > loaded?  
> >> >> >>>> You might be able to just build the kernel with 'nooptions 
> > NEW_PCIB'.
> >> >> >>>
> >> >> >>> I believe that I've got a similar problem with amdsbwd(4).
> >> >> >>> It needs some resources (I/O ports) that belong to ACPI.
> >> >> >>> The problem is that the driver attaches to isa bus which is under
> >> >> >>> isab->pci->pcib and those particular resources are not assigned to 
> > the Host-PCI
> >> >> >>> bridge.
> >> >> >>>
> >> >> >>> I think that you already made a suggestion that perhaps isa bus 
> > should  directly
> >> >> >>> attach to acpi bus when acpi is available.  Not sure if there are 
any
> >> >> >>> alternative approaches.
> >> >> >> 
> >> >> >> Can you try this:
> >> >> >
> >> >> > Not so much. :) the first and last patches I can apply to HEAD by 
hand,
> >> >> > but /sys/dev/acpica/acpi_pcib_acpi.c is only 387 lines long, so I'm 
not
> >> >> > even sure where to start.
> >> >> 
> >> >>   $ svn cat 
> > svn://svn.freebsd.org/base/head/sys/dev/acpica/acpi_pcib_acpi.c | wc -l
> >> >>   531
> >> >> 
> >> >> No difference here on ICH9, ichwd(4) still doesn't attach.
> >> >
> >> > Can you add some printfs to see if the new method is being called in
> >> > acpi_pcib_alloc_resource() and if it is failing when it is called?
> >> 
> >> rman_reserve_resource() fails with SYS_RES_MEMORY for isab0 when ichwd0
> >> tries to attach. And acpi_alloc_sysres() is not called when isab0 or
> >> isa0 are attached.
> >> 
> >>   isab0: found ICH9 or equivalent chipset: Intel ICH9 watchdog timer
> >>   ichwd0: <Intel ICH9 watchdog timer> on isa0
> >>   isab0: found ICH9 or equivalent chipset: Intel ICH9 watchdog timer
> >>   acpi_alloc_resource::acpi_alloc_sysres(child->nameunit=ichwd0, 
> > type=SYS_RES_IOPORT, *rid=0, start=1072, end=1079, count=8, flags=6) 
> > res=0xfffffe0007fed380 RF_ACTIVE activated
> >>   pcib0: allocated type 4 (0x430-0x437) for rid 0 of ichwd0
> >>   acpi_alloc_resource::acpi_alloc_sysres(child->nameunit=ichwd0, 
> > type=SYS_RES_IOPORT, *rid=1, start=1120, end=1151, count=32, flags=6) 
> > res=0xfffffe0007fed400 RF_ACTIVE activated
> >>   pcib0: allocated type 4 (0x460-0x47f) for rid 1 of ichwd0
> >>   acpi_pcib_acpi_alloc_resource::acpi_alloc_sysres(child->nameunit=isab0, 
> > type=SYS_RES_MEMORY, *rid=0, start=4275172368, end=4275172371, count=4, 
> > flags=6) res=0 failed to reserve
> >>   ichwd0: unable to reserve GCS registers
> >
> > Hmm, so it is called, it just fails when it is called.  The range is 
> > 0xfed1f410 - 0xfed1f413.  Can you verify that that is in the 'ACPI memory 
I/O' 
> > range in devinfo -u output?
> 
> It doesn't seem to be there (minus lines = hostres disabled).
> 
>    acpi0
>        I/O ports:
>            0x400-0x4bf
>            0x4d0-0x4d1
>        I/O memory addresses:
>            0xfed10000-0xfed1dfff
>            0xfed20000-0xfed8ffff
>      pcib0
>        pci0
>          isab0
>   -          I/O memory addresses:
>   -              0xfed1f410-0xfed1f413
>            isa0
>   -          ichwd0
>   -              ACPI I/O ports:
>   -                  0x430-0x437
>   -                  0x460-0x47f

Hmmm, so your BIOS just outright lies then? :(

Can you check devinfo -u with 'debug.acpi.disabled=hostres' to see where
those memory addresses show up?

-- 
John Baldwin